
Senior Applied ML Engineer | Kimchi (Model Router)
Own the end-to-end routing model: features, objective, training, calibration, evaluation, and rollout (shadow, A/B, full). Work with production fleet data, benchmark harnesses, and platform engineers to instrument per-request signals (including decode speed) and ensure production-parity with rigorous evals and error bars. Tech stack highlights: Python, PyTorch, SQL at scale (ClickHouse), OpenAI/Anthropic-compatible LLM APIs, and GitLab CI.




